People understand utterances spoken by many different people quite easily. However, complex mental processes underlie this ability. This proposal is concerned with specifying the mental processes and structures used to perceive speech and human voice. The experiments are designed to answer questions about the perceptual processes that encode voice, the attentional demands of processing voice, and the manner in which voice information is represented in human memory. The experimental methods consist of procedures that assess subjects' performance in tasks requiring them to listen to speech stimuli and make manual responses. Natural and synthetic speech are used in experiments involving perceptual identification, phoneme monitoring, auditory word recognition, memory recall, voice discrimination, and voice classification tasks. Many of the experiments make use of stimuli produced by a variety of male and female talkers, in sharp contrast to most of the research in speech perception that uses stimuli from a single talker. The results from these studies will provide empirical data relevant to developing a theoretical model of speech perception that includes a mechanism that encodes human voice. The methods used in these studies can be applied to different age groups and certain populations that exhibit speech-related perceptual disorders. The data obtained in this proposal can be used to establish normative baselines against which theories about development and perceptual disorders can be evaluated.